DESCRIPTION: "Adaptive" mutations occur in the apparent absence of cell division, after cells are exposed to non-lethal genetic selection, and have been detected only in genes whose function was selected. Their discovery challenges established ideas about the mechanisms of mutation and evolution. In one system, reversion of a lac frameshift in E. coli, aspects of the molecular mechanism are becoming apparent, and indicate a novel mutagenic route involving DNA double-strand breaks (DSBs), homologous recombination, DNA synthesis with polymerase errors, and possible suspension of post-synthesis mismatch repair. This proposal is aimed at beginning to test the generality of this mechanism elsewhere. The following areas will be investigated: 1. DNA Synthesis. 2. Mismatch repair be tested. We will test whether the repair is non-functional due to (i) undermethylation. 3. Recombination: The following will be tested: (i) various origins of the DNA partner used in recombination; (ii) whether whole recombination reactions, or just initiations without resolution of Holliday junctions, are required for adaptive mutation, iv) the ability of alternative (non-RecBCD) recombination systems of E. coli to support adaptive mutation. (v) A partial requirement found for RecF, but no other component of the RecF recombination pathway, will be investigated. 4. Origin of DSBs: Further, independent, evidence for the presence of DSB during adaptive mutation will be sought. Tests of several hypotheses for their origin will be made. 5. Generality of the system: The following will be investigated: (i) whether the F'-borne frameshift mutation reverts similarly in different chromosomal locations; (ii) whether the peculiarities of the mutation spectrum obtained when reverting a + 1 frameshift are found with other frameshifts; and (iii) whether the mechanism emerging in the frameshift system is generalizable to non-frameshift mutations in lac in the F' and the chromosome. Understanding the molecular mechanism will reveal whether the mutations represent Lamarkian-or Darwinian evolution, and will also provide a valuable new model for mutagenesis in non-dividing cells.